Microbial community assembly and hydrologic variability as controls on potential denitrification rates in restored freshwater wetlands

Denitrification is a microbial process in wetlands that transforms nitrate pollutants into nitrogen gas under anaerobic conditions. Restored wetlands do not always reach equivalent denitrification rates as those they were intended to replace, and this is sometimes linked to differences in microbial composition. Further, increased drought and variability in flood regimes represents an additional threat because the stress of fluctuation between extreme dry or saturated conditions could alter the microbial community itself and the ability of remaining microbial taxa to transform nitrate via denitrification. One of the goals of this dissertation was to identify either microbial or hydrologic factors that could limit the ability of restored wetland communities to reach high potential denitrification rates. Based on a survey of 30 restored and 15 reference wetlands, restored wetlands surprisingly reached equivalent rates as observed from reference sites. This survey was conducted twice: Once during a drought in 2012, and again in 2013 following unusually intense floods. Similar results were found both years, but average rates were an order of magnitude greater in 2013 than in 2012. These potential denitrification assays were performed under identical saturated anaerobic conditions in the lab each year, so the differences observed must be due to inherent differences in the active microbial community. The magnitude increase in rates between the two years could be explained by soil variables such as pH and moisture, as well as to the abundance of nirS-harboring denitrifiers in the community. Soil moisture and oxygen availability strongly influence microbial community composition, so regions of a single floodplain wetland may contain vastly different communities depending on the hydrology. Another goal of this dissertation was to determine if past hydrologic variability filters microbial taxa such that these communities exhibit high resistance or resilience and high functional stability in the face of altered hydrology. I established an experiment where I manipulated the hydrology of soil collected from upland, lowland and the transition region of a single wetland. The transition region contained a unique composition compared to the upland and lowland regions, and composition remained relatively stable compared to the other regions, indicating either high resistance or resilience. This community also exhibited stable potential rates compared to the other regions. Since denitrification can also be influenced by soil factors, such as pH and texture, which differ among wetlands and even differ among regions of the same wetland, my final study included an experimental design that removed any differences between soil types. Sterilized soil mesocosms were inoculated by live wetland soil collected from wetlands that exhibited different historical flood regimes: Flashy floods, long extended floods, and high variability from year to year. These wetlands exhibited distinct microbial composition initially. Following inoculation of the sterile mesocosms, I manipulated the hydrology as I did in the previous study, but this time I measured potential denitrification at multiple time points, rather than only at the beginning and end. The manipulated hydrology was the strongest driver of rates, but the mesocosms inoculated with communities from the wetlands with high variability from year to year consistently showed lower rates than those inoculated by communities from the other wetlands. Increased drought followed by intense floods due to climate change will influence denitrification function in wetlands. Restored wetlands can reach equivalent denitrification rates as those that they were intended to replicate, but both restored and reference wetlands are now threatened by increased drought and hydrologic variability.U of I OnlyAuthor requested U of Illinois access only (OA after 2yrs) in Vireo ETD syste